Kapler re-signs in Tampa and further proof Fangraphs is ridiculous

[thebruce44]

Looper had some half way decent statistics this year

 

Other than wins (slightly useless "stat"), what was decent about Looper? I don't need Fangraphs to tell me that he was a bottom scraping SP in the league in 2009, most of his stats will tell you that.

 

Yea, he was pretty bad all around. I was thinking of a discussion I had with some people during the season (not on this board) about how Parra had actually been much better than Looper at that point in the season, as shown in nearly every statistic besides ERA and Wins. Looking back on the season stats now though, Looper was just god awful. Maybe someone like Carlos Villinueva before this year is a better example, or Soup when he was in his final year with the Cardinals. ERA is a flawed statistic in its own right.

 

The point was that with most statistics you can find plenty of examples that don't pass the stink test just like the Morgan vs Braun issue. That isn't a good reason in my mind to say Fangraphs or WAR is ridiculous though.

[brewjihad]

Calculating how many wins above replacement level a player is isn't nearly as complicated as some of you might think. If you can tell me a player's wOBA (I hope everyone knows what wOBA is, if you don't it gives the average amount of runs produced per event, like homer, single, out, walk, etc compared to an average situation), UZR (which maybe isn't a very good indicator of how good of a defender a player is, but is a good indicator of how good of a defender the player was in the past year), amount of plate appearances, and position, it's not too hard to figure out their WAR. Heck, if you hate UZR for some reason, you can use any measure of how many runs above/below average the defender was compared to his positional peers. I'll do it by hand here quick to show that it's pretty simple. Mike Cameron is my example. I only used fangraphs to check the UZR and wOBA, I didn't look at the value section to show you I can do it by hand. If you really wanted me to I could do wOBA by hand if I knew the following things: PAs, HRs, 3B, 2B, 1B, BB.

 

wOBA: .346

Plate Appearances: 628

UZR: 9.9

Position: CF, so adjustment of 2.5 for a full season

 

Okay, so to convert wOBA to runs you subtract league average from Cam's wOBA. Average is .333 so we end up with .013. Divide that by 1.15 and multiply by plate appearances. That's ((.346-.333)/1.15 * 628) I get 7.1 runs above average. Add the UZR+Adjustment, which I'll give 2.5 assuming close to a full season of pt in center. So, finally, the replacement level adjustment. All of the measures so far compare to average, so we want to add a bonus for taking time that could've been played by a replacement level player. A replacement level is 20 runs below average over 600 plate appearances. Because it's cumulative, a player who racks up 600 PAs gets 20 runs. So an exactly average player in every other measure gets +20, which is defined as about average (a 2-win player). For Cam, he's over 20, He gets the 20 runs for 600 then I divided 28/600 and multiplied that by 20 again, which comes out to only about .9. So he gets 20.9 above replacement level.

 

So to sum that up: 7.1 + 2.5 + 9.9 + 20.9 = 40.4 runs above a replacement level player. That converts to about 4.04 wins.

 

Now I checked fangraphs, and their value says 43.0 runs above replacement, and 4.3 wins. So I was pretty close. Looks like the hitting is 2 runs different, don't know what would cause that. But the point is made.

 

*(The positional adjustments were made by Tango and were formed by comparing players who moved positions over the years. This accounts for the fact that better defenders play center field than, say, left, so the average defender is better and provides more value.)

 

*(Also, about defense. People tend to think that defense stats aren't reliable because you might see a -10, then a +5, and then a -5, and it doesn't seem like a stat that volatile could be reliable. It's a trick, kinda. First, you need to think of a full season of defense as about 200 plate appearances of offense. Then, remember that it's compared to average. So in comparison you could have a hitter with 3 sections of 200 PAs with wOBAs of .320, .360, and .340. Convert that to runs compared to average and it's about -5, +5, and +1. The way defensive stats are presented, as , makes them seem less reliable.)

[endaround]

CF and LF are different positions and are treated as such by UZR. Don't know why you think they are pooled.

[Hate those Rodents]

Most of us understand perfectly well how WAR is calculated and how they arrive at the average $ figure per win and simply don't agree with the methodology involved.

 

You are one of the posters who's been willing to do his homework and actually learn about this stuff. I don't agree with all your criticisms (some certainly point out significant weaknesses) but at least you've armed yourself with much of the knowledge needed to carry on an intelligent discussion about the topic. Despite your claim, most of the detractors posting here have no idea how it's calculated. They don't agree with the conclusion, so that's the end of that. Hell, at least one poster still seems convinced that the methodology is a secret.

I will admit to being what you are describing here. I guess this is the abortion issue for me. I still feel I am very knowledgeable on baseball and have a lot to contribute to the board, but I am not going to spend time learning all sorts of formulas that only show up on the most hardcore of baseball sites, especially when the end result the formula yields is so wrong in my eyes. If there is an obscure formula that comes closer to calculating what a player is worth the previous season or what they should be paid the following season, I'd personally be all about studying that particular formula.

[thebruce44]

I am going to try to make an analogy that may be way off and not make sense to some people, but it is how I view WAR and similar statistics/measures.

 

In my line of work, designing interstate highways, there are often several valid layouts for a roadway. Often several designs will be presented with various plus and minuses to each. I often have one design I prefer right off the bat. Someone else in my office will sometimes prefer another design, and a third person at the DOT may prefer the third one. In other instances, one design will be obviously the best choice to the majority of people and the decision is fairly easy.

 

One way to discuss the designs is to make a decision matrix. Without explaining the entire methodology it is at its core a list of weighted plusses and minuses for each option. In the end, you are able to arrive at a final score for each option. Over time, this methodology can be adjusted so that when one option is obvious the matrix hopefully arrives at that being the best solution. The decision matrix is never the end all, be all. It is an attempt to quantify the decision process since it can otherwise be difficult to compare a bunch of subjective opinions.

 

I view WAR similar to a decision matrix. It isn't perfect and hopefully over time adjustments will be made to refine it. Right now, it is an attempt to quantify the "who is better than who" opinions we all have and to show just how much better one player is than another. It is not the end all, be all, but it is a great first step in those types of discussions and the best "objective" resource currently available for well rounded player comparisons.

 

I don't like when I see people say something along the lines of "I don't need WAR to tell me player A is better than player B, I just know." That is your opinion and is perfect valid, but why knock an attempt to quantify a range of opinions when better options are not available?

 

Ideally we would just put two players side by side and compare their OPS, but that doesn't give us the whole picture when you factor in position or defense. You may have your opinion and prefer to stick to it, but unless you are able to watch every MLB game to effectively judge defensive value while simultaneously coming up with your own way to compare different positions, I think an attempt to use statistics the way WAR does is just as valuable.

 

Furthermore, in reading this thread you will notice that most of the people who understand WAR and its value have pointed out that it is valuable but not perfect. Most of the people who seem to not like it look at a few end results they disagree with or even the player value and decide the whole idea is ridiculous. There are plenty of posters in between and the "closed minded" comment isn't directed at them.

 

All that I am asking is that you do a little more research than "Kapler was supposed to be worth this much" or "Morgan isn't more valuable than Braun in my opinion" before you arrive at the conclusion that fangraphs is ridiculous.

[Ennder]

Looper was more or less the exact same pitcher he has always been, he just had more balls go over the fence this year so his ERA jumped. If he really were significantly different(like if he was leaving them up in the zone more etc) he'd have more flyballs in general but his rates were exactly like normal. Just a few more went over this year. This kind of thing happens all the time and 90% of the time it just corrects itself the next season.

 

His xFIP pretty much tells the story.

 

2009 - 4.90

2008 - 4.59

2007 - 4.93

2006 - 4.33

2005 - 4.85

 

Back in 2004 he was a better pitcher because he induced a lot more groundballs but from 2005 on he has been somewhere between a 4.50 and 5.00 ERA true talent pitcher. With STL since they had a pitchers park and good defense his ERA was bit lower and obviously the years as a RP tend to have lower ERA as well.

 

As for WAR, I think it is pretty much useless for evaluating talent of pitching, especially for a single season. On the positional side of things I like it a bit more but the defensive metrics need more than a single season of data as well. I'd certainly rather compare players using WAR than just using fan perception though. It is very plausible that an extremely good defensive CF who is a mediocre hitter could be more valuable than Braun, fans tend to overrate SLG and underrate defense in general. I'm not sure I buy that Morgan is that good defensively though.

[joepepsi]

In 2003, I was spouting off on this very board that I didn't think the difference between good and bad defenders was large enough to make good defense very valuable. I had no evidence of course; I just KNEW it was right. How pathetically wrong I was.

The difference between good and bad defense is significant. The difference between good and average, not so much. It's laughable to me that Braun's defense is 15 runs worse than a replacement level player's defense. What is even harder for me to grasp is the notion that Nyjer's defense is so great that's it's equal to Braun's offensive contributions. If it were true that great defense was equal to great offense than a team of Neifi Perez's would win the World Series on a regular basis.

[logan82]

A replacement player is usually though of as an average player on defense. Braun isn't a good defender and I am surprised that it is so hard to believe that he was 15 runs below average on defense.

[Invader3K]

He might not be "good", but he isn't as bad as he is made out to be on this forum and others. Look how many butchers are playing LF around the league. Braun is pretty decent in comparison. I'm not saying he's a gold glove candidate, but his defense is really more maligned than it reasonably should be at this point.

[sheethead]

I don't like when I see people say something along the lines of "I don't need WAR to tell me player A is better than player B, I just know." That is your opinion and is perfect valid, but why knock an attempt to quantify a range of opinions when better options are not available?

 

Who says they "just know?" There are a plethora of actual statistics to look at when judging players. I simply am not in tune with having to multiply by 2.5, put to the power of the wind/continental divide, doubled and then taken as the reciprocal to end up at the conclusion that Nyjer Morgan is more valuable than Ryan Braun. Its comical to me. I get the attempt, I am just saying there is no way I could buy in at any kind of level to where I would be citing the results in an attempt to validate my or anyone else's opinion.

 

but unless you are able to watch every MLB game to effectively judge defensive value while simultaneously coming up with your own way to compare different positions

I don't think you have to watch every game to determine the value of a players defense, unless you happen to watch Luis Castillo in that one game (then you might have to watch a few extra Mets games). I am also not sure why we have to compare 1B to Catchers to CFers. Every team needs at least one of each, even if one is -1.5 time less valuable than the other.

Now I checked fangraphs, and their value says 43.0 runs above replacement, and 4.3 wins. So I was pretty close. Looks like the hitting is 2 runs different, don't know what would cause that. But the point is made.

 

I would think the point is that the numbers should be the same.

 

The positional adjustments were made by Tango and were formed by comparing players who moved positions over the years.

 

This is where judgment kicks in, and quantification checks out (and may lead to funky end results).

[endaround]

Yeah all thinking like that got us was a 1 year turnaround of the Mariners!

 

Now Morgan has issues in that anytime you go over 20 runs above average on defense you're questionable. So if you want to knock off a few runs, feel free. The point is that the difference between Braun and Morgan is about 30-40 runs on defense.

[sheethead]

Yeah all thinking like that got us was a 1 year turnaround of the Mariners!

 

I am not sure I would go as far as saying the Mariners are turned around.

[brewjihad]

I simply am not in tune with having to multiply by 2.5, put to the power of the wind/continental divide, doubled and then taken as the reciprocal to end up at the conclusion that Nyjer Morgan is more valuable than Ryan Braun. Its comical to me.

 

What if you just divide the difference between wOBA and .333 by 1.15, multiply by plate appearances, then add that to defense, position, and +20 runs per 600 plate appearances. That's what the WAR formula is. It's not too complicated, it's like a middle school math formula.

 

I would think the point is that the numbers should be the same.

 

I'm showing you that Fangraphs doesn't have a magical WAR machine that spits out the values. I can do it entirely by hand.

[sheethead]

I'm showing you that Fangraphs doesn't have a magical WAR machine that spits out the values.

 

I don't have a problem understanding this. Please stop insinuating that I do.

 

I can do it entirely by hand.

 

It doesn't match. I will give you credit for the work, but you are losing points for having a different answer. Middle school math is a science, and there is always a single answer. This is why I claim they are using figures that are not publicized. Unlike how people keep incorrectly stating I think the methodology is proprietary, its what goes into the actual calculation that remains a mystery. Otherwise, we would all get the same numbers, right? I guess close isn't good enough for me to say you are dead on.

[stalton]

In reading this thread, I have an image of Jackie Chan "singing" Edwin Starr....(Even though I don't believe that its good for absolutely nothing) I've always been a fan of fangraphs (haha) but as far as WAR is concerned, I think it does slightly overvalue defense, but thats just their opinion versus mine. Like thebruce, I've made quite a few decision matrices myself, and I think the people at FG really put alot of weight into defense.

 

In looking at comparing WAR for Center Fielders:

3 of the top 4 CF WAR's have a higher defensive Runs Above Average then offensive Runs Above Average...to me that states that the best CF'ers are gold glovers, unless they can really mash the ball ( Kemp is the outlier of the top 4; Frank Gutierrez, Morgan and Cameron are the defensive specialists.)

 

According to Fangraphs, Cam is the 4th Best CF'er in regards to WAR.....Regardless of what that number is, How many people would agree that there were only 3 other better CF'ers overall?

[logan82]

How many people would agree that there were only 3 other better CF'ers overall?

 

I would. Cam plays good defense and has a really good bat for a CF. I don't think saying Cameron is a top 5 CF is a stretch at all.

[joepepsi]

How many people would agree that there were only 3 other better CF'ers overall?

 

I would. Cam plays good defense and has a really good bat for a CF. I don't think saying Cameron is a top 5 CF is a stretch at all.

I don't either, and going strictly by OPS you'd find that Cameron is 6th among outfielders who played at least 100 games in CF. His OPS was .795, .001 behind Fukdome, who isn't nearly as good as good a defender and started only 105 games in CF. Cam started 146 games in CF, so it's safe to say he's clearly a more valuable player than Fukudome. It's not a stretch to put Cameron in the top 5 CF'ers in MLB; I just don't think you need fangraphs to tell you that when handier, simpler stats already make the case.

 

On another topic, if a replacement level player is considered to be an average defender, then the value of a good defender is less than I even thought. If any Schmoe is average out there, it's not worth paying extra for somebody really good. It's really difficult for me to believe Braun cost 15 more runs in LF than an average LF'er. He'd have to be booting balls, missing flies, making fewer attempts than a replacement player would out in LF. Where's the data that shows he was missing so many plays throughout the year that opposing teams scored 15 more runs than they otherwise would have?

[sheethead]

If you are above average and have had 600 PAs, you get 20 bonus points. Is that right?

[brewjihad]

I don't have a problem understanding this. Please stop insinuating that I do.

 

By mocking the compilation of stats by saying this, you are making it seem like you don't understand it:

 

I simply am not in tune with having to multiply by 2.5, put to the power of the wind/continental divide, doubled and then taken as the reciprocal

 

On to this:

 

It doesn't match. I will give you credit for the work, but you are losing points for having a different answer. Middle school math is a science, and there is always a single answer. This is why I claim they are using figures that are not publicized. Unlike how people keep incorrectly stating I think the methodology is proprietary, its what goes into the actual calculation that remains a mystery. Otherwise, we would all get the same numbers, right? I guess close isn't good enough for me to say you are dead on.

 

If you go through my post and compare the numbers, the only discrepancy is in the offensive runs above average column. It's a 2-run difference, and I honestly don't know what would cause it. The conversion from wOBA to batting runs is difference from league average / 1.15 * plate appearances. I don't know why you're saying the numbers aren't publicized, every number in the WAR calculation is on the player's page.

 

If you are above average and have had 600 PAs, you get 20 bonus points. Is that right?

 

Again, you're showing that you don't quite understand exactly what we're measuring. Defense and offense are both compared to a league average player. If you added the offense, defense, and position adjustment the final measure would be WAA, wins above average, but there's 2 problems with that. First of all, a player that gets 100 plate appearances and hits well enough to be +3 runs on offense and average on defense is much less valuable than a player who gets 600 plate appearances and hits -5 runs below average and plays average defense. Secondly, average players are not available for free, and replacement level players are. Comparing a player's production to replacement level provides better context for his performance. So think of it this way: an exactly average hitter with a wOBA of .333 and defense+adjustment of exactly average who accumalted 600 plate appearances would be 0 wins above average. But he's 2 wins or 20 runs better than a replacement level player. If he's 10 runs below average total, he's still a 1-win player. If he's 4 wins above average, he's a 6-win player. If that same initial hitter-- .333 wOBA, average defense-- only got 300 plate appearances, he'd get +10 runs, so he's a 1-win player too. This is why WAR is cumulative and allows you to compare every player to each other.

[kramnoj]

I don't think you have to watch every game to determine the value of a players defense,

 

If you want to accurately measure how valuable any single player was in any given season, you absolutely have to watch every single game. Otherwise you are just making an observation about their talent, and not their actual contribution. A single season of UZR should never be used to definitely say how good a player is, but it can tell you how that player measured in a single year. That is all it should be used for. Once we have enough seasons, we can use that combined with scouting to have a clearer, more confident description of their ability.

[logan82]

An outfielder doesn't have to boot balls to be a bad fielder. Just not getting to as many balls would be enough. If you look at the breakdown of Braun's UZR you see that it rates him very poorly on range and very well(in comparison to other players) on the error portion. 15 runs may be high, but if you ever look at offensive stats really closely, an extra hit a week adds up to about 85 points of OPS. I don't think it would be much of a stretch to say that Braun lets an extra hit fall in front of him or get over his head once a week. An extra ball falling in here and there adds up to quite a big deal over an entire season.

 

I guess my point is that if a seemingly small event(one hit every week) on offense can lead to such a drastic difference(85 points of OPS) over the course of a season, a seemingly small difference in quality of defense would probably show up as a pretty large difference in defense as well.

[Oldcity]

Yes, pretty much. Because his defense really is not all that bad anymore and his bat is so darn good, that's why.

 

Braun's defense is still horrible.

[Invader3K]

Braun's defense is still horrible.

We'll just have to agree to disagree on that then. I think "horrible" is a gross overstatement. I would say he's mediocre to average for an NL left fielder.

[sheethead]

It's a 2-run difference, and I honestly don't know what would cause it.

 

That's what I'm saying.

[rluzinski]

5 engineers at my work could work on the same peoblem and all get slightly different answers. According to you, that invalidates all the methodologies used. That's a weak argument.

 

Fangraphs tweaked Tom Tango's methodology and will come up with a slightly different number than him. They had something like a 7 part series outlining exactly how hey calculate it. Others are free to tweak, refine and improve as well. It's a model. If somone is arguing that one player was worth 2 runs/.2 wins more than fangraphs indicates, they missing the point.

 

As for defense, every missed play is worth about -0.8 runs (rule of thumb). that's the combined value of the extra basrunner, the missed out and the extra batter. So -15 runs equals about 19 less plays made over 162 games. One less play every 8-9 games. And we are really talking about probabilities of making the play. If Braun missed a ball that's converted into an out 50% of the time, he's debted .8 runs x .5 = .4 runs.

 

I don't think Braun really is that bad of a defender but he may have just had a bad year. It's also possible that UZR didn't estimate the value of his performance very well for a season.